Evaluating user-defined conditions

[Chris Jack]

Can I suggest you consider including some rudimentary idea of "cost" when you're deciding whether to allow the query to run or not. "Cost" could be in terms of anticipated rows returned and/or total anticipated CPU time.
This could be a slippery slope as to do it well you'd have to start creating histograms of your data.
If someone has any time (ha!): a nice addition to some of the modules mentioned would be to include some sort of self-monitoring: so if memory goes up more than a certain amount of CPU time goes over some threshold: the module decides to abort itself.
CheersChris